Text is Data: Making Use of Archived NOAA Text Products

When thinking about data, we often think about numbers and observations. However, text can also be data. Forecasters at the National Weather Service do an excellent job creating accurate text-based forecasts every day, but this hard work is often overlooked once the next forecast is issued. Until now, it was very challenging to make use of all these data. I developed a new tool called the Forecast Search Wizard that makes it easier to analyze archived NOAA text products to obtain valuable information. The Forecast Search Wizard is a program that allows researchers to search keywords in NOAA/NWS text products (e.g., Area Forecast Discussions) to identify the occurrence of case studies, particular weather events, or other specific information. This tool will not only save the researcher time, but it will also improve the depth of the research.

In many meteorological research projects, one of the first steps is to identify case studies. The researcher typically needs to gather and analyze numerical data to find events that share common characteristics, or the researcher needs to know when similar events happened. The former often requires a lot of time that may result in numerous false flags and missed cases, while the latter requires that the researcher know every occurrence of these events. It is difficult to know every occurrence and so the latter is likely to result in a limited number of cases. The Forecast Search Wizard has been optimized to quickly and thoroughly search for keywords. It is a quick way to find a more accurate list of potential case studies, with fewer false flags and fewer missed cases. For example, I was able to identify 438 days where Cold Air Damming (CAD) was mentioned in the forecast by the National Weather Service for the Washington/Baltimore area (2007-Present). This search took just moments. The possibilities are endless.

I will demonstrate how the Forecast Search Wizard can be a useful tool, explain its capabilities, and discuss its limitations. This robust program can quickly and accurately identify keywords in archived forecasts to make this information more accessible. The Forecast Search Wizard will provide the user valuable data that can improve the quality of their research. This tool will be made available to the public.